Federated Learning (FL) is a promising distributed model training paradigm without the leakage of local data. In wireless communication networks, energy-constrained mobile devices may encounter outage when uploading local gradients to edge server due to fading channels, leading to lost devices and thus performance degradation. Retransmission is a classical mechanism to address the outage issue, which however increases energy consumption and delay of mobile devices. To this end, we propose an adaptive retransmission strategy for wireless FL, where the server decides whether to request retransmission according to the number of transmission outage, and devices decide whether to respond based on their channel conditions and remaining energy. A convergence analysis is conducted for the proposed FL framework. Based on this, we aim to minimize the convergence error by jointly optimizing the retransmission decisions and the allocation of retransmission slots, under the devices’ long-term energy budget and the communication resource constraints. Since the decisions are coupled across different iterations, we design a Lyapunov-based online algorithm to solve this problem. Through experiments, we find that the effect of outage is trivial sometimes and retransmission is not always necessary in FL because channel conditions and data distributions affect the degree of the retransmission gain, and that the retransmission strategy benefits wireless FL.

Split Federated Learning (SFL) splits and collaboratively trains a shared model between clients and server, where clients transmit activations and client-side models to server for updates. Recent SFL studies assume synchronous transmission of activations and client-side models from clients to server. However, due to significant variations in computational and communication capabilities among clients, activations and client-side models arrive at server asynchronously. The delay caused by asynchrony significantly degrades the performance of SFL. To address this issue, we consider an asynchronous SFL framework, where an activation buffer and a model buffer are embedded on the server to manage the asynchronously transmitted activations and client-side models, respectively. Furthermore, as asynchronous activation transmissions cause the buffer to frequently receive activations from resource-rich clients, leading to biased updates of the server-side model, we propose Generative activations-aided Asynchronous SFL (GAS). In GAS, the server maintains an activation distribution for each label based on received activations and generates activations from these distributions according to the degree of bias. These generative activations are then used to assist in updating the server-side model, ensuring more accurate updates. We derive a tighter convergence bound, and our experiments demonstrate the effectiveness of the proposed method.

Federated Learning (FL) has gained significant attention as a distributed machine learning paradigm, leveraging edge clients for collaborative training while preserving data privacy. However, statistical heterogeneity, arising from diverse client data distributions, poses a significant challenge to FL performance. In this work, we propose a novel heterogeneity-aware client selection scheme based on prototype margin to mitigate these challenges. By utilizing prototype learning, we capture effective data representations, allowing client selection to better account for both data quality and distribution. The proposed scheme is designed for practical wireless communication networks, where clients are clustered and we select participants based on joint prototype margin and wireless channel. To further optimize the training process, we jointly optimize client selection and bandwidth allocation, accounting for dynamic network conditions and heterogeneous client capabilities. Experimental results demonstrate that our approach significantly improves model accuracy and reduces communication overhead.

Over-the-air computation (AirComp) exploits the waveform superposition of wireless channels for fast data aggregation from multiple devices. The implementation of AirComp requires amplitude alignment among devices, which requires better channel conditions. Meanwhile, movable antenna (MA) is an emerging method to create better channel states via local antenna movement. To fully utilize the channel gain obtained by adjusting the positions of MA, we consider a MA-enabled AirComp system equipped with one-dimensional MA at transmitter to aggregate wireless data from a large number of devices. We aim to minimize the mean squared error (MSE) by jointly optimizing the antenna position vectors (APV), transmit power, and denoising factor at devices. To address this highly non-convex problem, an alternating optimization (AO) based algorithm is adopted by decomposing it into two sub-problems for power control and APV optimization, respectively. Specifically, we obtain a semi-closed form solution of transmit power control and denoising factor under any given antenna position, and then use second-order Taylor expansion to derive a more tractable MSE counterpart for APV optimization under successive convex approximation (SCA) technique. Experimental results show that, compared with other benchmark schemes, our proposed scheme demonstrates better performance.

Split Federated Learning (SFL) is a distributed machine learning framework where the models are split and trained on the server and clients. However, data heterogeneity and partial client participation result in label distribution skew, which severely degrades learning performance. To address this issue, we propose Concatenated Activations Enabled SFL with Logit Adjustments, in which activations from client-side models are concatenated as the input of the server-side model to centrally adjust label distribution across different clients, and logit adjustments in the loss functions of both server-side and client-side models are performed to deal with the label distribution variation across different subsets of participating clients. Experiments demonstrate the superiority of the proposed method compared with the traditional schemes.

This paper focuses on the multi-splitting of DNN over wireless multi-hop networks to distribute the computing over multiple network nodes for achieving efficient edge infer-ence. The challenge is how to choose an inference routing in which both the transmission and inference can be efficiently relayed hop-by-hop. We propose a DNN multi-splitting method along with dynamical early-exit of inference based on communication and computation conditions. We formulate and then solve an optimization problem of joint routing, split points selection, and model deployment to minimize the end - to-end inference latency. The experimental results demonstrate the superiority of our work in reducing inference latency.

Federated learning (FL) is a promising distributed machine learning paradigm to train a machine learning model without the leakage of local data. However, as the sizes of models are increasing while Internet of Things (IoT) devices are heterogeneous and capability-limited, FL faces performance bottleneck. In this article, we propose a cooperative device-to-device (D2D)-based partial training scheme for wireless FL. Specifically, the IoT devices in each cluster extract and train the nonoverlapping submodels from the global model, and the trained submodels are transmitted to the cluster head (CH) to form a whole local model via D2D links. Then the CHs upload the local models to the server for global aggregation. We first conduct the convergence analysis for the proposed wireless FL scheme. Then a joint optimization problem is formulated to minimize the average delay by the optimization of model division, device selection, and bandwidth allocation. An efficient algorithm is proposed to solve this nonconvex problem. Comprehensive experiments verify the efficiency of the proposed scheme.

Federated learning (FL) enables distributed devices to train a shared machine learning (ML) model collaboratively while protecting their data privacy. However, the resource-limited mobile devices suffer from intensive computation-and-communication costs of model parameters. In this paper, we observe the phenomenon that the model parameters tend to be stabilized long before convergence during training process. Based on this observation, we propose a two-timescale FL framework by joint optimization of freezing stabilized parameters and controlling transmit power for the unstable parameters to balance the energy consumption and convergence. First, we analyze the impact of model parameter freezing and unreliable transmission on the convergence rate. Next, we formulate a two-timescale optimization problem of parameter freezing percentage and transmit power to minimize the model convergence error subject to the energy budget. To solve this problem, we decompose it into parallel sub-problems and decompose each sub-problem into two different timescales problems using the Lyapunov optimization method. The optimal parameter freezing and power control strategies are derived in an online fashion. Experimental results demonstrate the superiority of the proposed scheme compared with the benchmark schemes.

The development of artificial intelligence (AI) provides opportunities for the promotion of deep neural network (DNN)-based applications. However, the large amount of parameters and computational complexity of DNN makes it difficult to deploy it on edge devices which are resource-constrained. An efficient method to address this challenge is model partition/splitting, in which DNN is divided into two parts which are deployed on device and server respectively for co-training or co-inference. In this paper, we consider a split federated learning (SFL) framework that combines the parallel model training mechanism of federated learning (FL) and the model splitting structure of split learning (SL). We consider a practical scenario of heterogeneous devices with individual split points of DNN. We formulate a joint problem of split point selection and bandwidth allocation to minimize the system latency. By using alternating optimization, we decompose the problem into two sub-problems and solve them optimally. Experiment results demonstrate the superiority of our work in latency reduction and accuracy improvement.

Federated learning (FL) is a promising distributed learning framework where distributed clients collaboratively train a machine learning model coordinated by a server. To tackle the stragglers issue in asynchronous FL, we consider that each client keeps local updates and probabilistically transmits the local model to the server at arbitrary times. We first derive the (approximate) expression for the convergence rate based on the probabilistic client selection. Then, an optimization problem is formulated to trade off the convergence rate of asynchronous FL and mobile energy consumption by joint probabilistic client selection and bandwidth allocation. We develop an iterative algorithm to solve the non-convex problem globally optimally. Experiments demonstrate the superiority of the proposed approach compared with the traditional schemes.

The development of artificial intelligence (AI) provides opportunities for the promotion of deep neural network (DNN)-based applications. However, the large amount of parameters and computational complexity of DNN makes it difficult to deploy it on edge devices which are resource-constrained. An efficient method to address this challenge is model partition/splitting, in which DNN is divided into two parts which are deployed on device and server respectively for co-training or co-inference. In this paper, we consider a split federated learning (SFL) framework that combines the parallel model training mechanism of federated learning (FL) and the model splitting structure of split learning (SL). We consider a practical scenario of heterogeneous devices with individual split points of DNN. We formulate a joint problem of split point selection and bandwidth allocation to minimize the system latency. By using alternating optimization, we decompose the problem into two sub-problems and solve them optimally. Experiment results demonstrate the superiority of our work in latency reduction and accuracy improvement.

The development of Industrial Internet of Things (IIoT) has completely changed the traditional manufacturing industry. The data exchange between controllers and actuators needs to achieve extremely low delay in IIoT. Due to the limited communication resources, it is necessary to reasonably schedule data flow to reduce delay. Although the studies of data flow scheduling exist in IIoT, they have not considered the impact of time-varying environmental factors and most of them adopted centralized scheduling schemes, which increase computation and communication cost rapidly in large-scale network scenarios. In this article, the consensus-based distributed optimal transport (OT) algorithm is proposed to optimize data flow scheduling for IIoT networks. Specifically, a data flow scheduling optimization mechanism based on time-varying environmental factors is proposed and an online distributed data flow scheduling optimization algorithm is designed. Compared with the random data flow scheduling algorithm, numerical results show that the proposed algorithm can maximally reduce the average delay by 87%, increase the transmission rate and the spectral efficiency by 157% and 98%, respectively.

Distributed Stein Variational Gradient Descent (DSVGD) is a non-parametric distributed learning framework for federated Bayesian learning, where multiple clients jointly train a machine learning model by communicating a number of non-random and interacting particles with the server. Since communication resources are limited, selecting the clients with most informative local learning updates can improve the model convergence and communication efficiency. In this paper, we propose two selection schemes for DSVGD based on Kernelized Stein Discrepancy (KSD) and Hilbert Inner Product (HIP). We derive the upper bound on the decrease of the global free energy per iteration for both schemes, which is then minimized to speed up the model convergence. We evaluate and compare our schemes with conventional schemes in terms of model accuracy, convergence speed, and stability using various learning tasks and datasets.

Mobile edge computing (MEC) is a promising tech-nology to support the latency-critical applications of mobile devices by offloading complex computation tasks to edge servers. However, the mobility of devices yield a great challenge on de-livering reliable continuous services, especially for those latency- critical applications. Motivated by the fact that the user's location changes slower than the task arrivals, we propose a two-timescale mobility management framework by joint service migration and power control. The management design is formulated as a long-term energy minimization problem, subject to the reliability requirement of the latency-critical application. Leveraging the Lyapunov optimization technique, we develop an online two- timescale control algorithm to solve the problem. The simulation results demonstrate that our proposed online algorithm can significantly improve the energy and reliability performance compared to the baselines.

Cognitive radio networks (CRNs) are expected to be promising techniques for improving the spectrum efficiency of wireless network utility in the squeezed sub-6-GHz frequency bands. Nevertheless, frequency allocation and transmission scheduling for secondary users (SUs) in CRNs suffer from no prior knowledge of other SUs’ network behaviors or the distribution of the amount of data generated at each SU. As a countermeasure, this article develops a protocol for the joint channel selection and transmission scheduling such that SUs with heterogeneous data transmission demands could be served with limited spectrum resources. Then, we formulate the dynamic optimization of the protocol as mutually embedded Markov decision processes (MDPs). To address the intractable MDPs,$Q$-learning-based channel selection and transmission scheduling based on reinforcement learning with basis function approximation are, respectively, proposed. It is shown that compared with various baselines, the proposed channel selection algorithm enables each SU to select the best frequency-domain channel that does not interfere with other SUs. In particular, the proposed transmission scheduling algorithm outperforms algorithms based on off-the-shelf approaches, such as$Q$-learning and Lyapunov optimization, in terms of both energy efficiency and long-term accumulative amount of bits at each SU.

With the development of Internet-of-Things (IoT), we witness the explosive growth in the number of devices with sensing, computing, and communication capabilities, along with a large amount of raw data generated at the network edge. Mobile (multi-access) edge computing (MEC), acquiring and processing data at network edge (like base station (BS)) via wireless links, has emerged as a promising technique for real-time applications. In this paper, we consider the scenario that multiple devices sense then offload data to an edge server/BS, and the offloading throughput maximization problems are studied by joint radio-and-computation resource allocation, based on time-division multiple access (TDMA) and non-orthogonal multiple access (NOMA) multiuser computation offloading. Particularly, we take the sequence of TDMA-based multiuser transmission/offloading into account. The studied problems are NP-hard and non-convex. A set of low-complexity algorithms are designed based on decomposition approach and exploration of valuable insights of problems. They are either optimal or can achieve close-to-optimal performance as shown by simulation. The comprehensive simulation results show that the sequence-optimized TDMA scheme achieves better throughput performance than the NOMA scheme, while the NOMA scheme is better under the assumptions of time-sharing strategy and the identical sensing capability of the devices.

Mobile edge computing (MEC) is a promising technology for enhancing the computation capacities and features of mobile users by offloading complex computation tasks to the edge servers. However, mobility poses great challenges on delivering reliable MEC service required for latency-critical applications. First, mobility management has to tackle the dynamics of both user’s location changes and task arrivals that vary in different timescales. Second, user mobility could induce service migration, leading to reliability loss due to the migration delay. In this paper, we propose a two-timescale mobility management framework by joint control of service migration and transmission power to address the above challenges. Specifically, the service migration operates at a large timescale to support user mobility in the multi-cell network, while the power control is performed at a small timescale for real-time task offloading. Their joint control is formulated as an optimization problem aiming at the long-term mobile energy minimization subject to the reliability requirement of computation offloading. To solve the problem, we propose a Lyapunov-based framework to decompose the problem into different timescales, based on which a low-complexity two-timescale online algorithm is developed by exploiting the problem structure. The proposed online algorithm is shown to be asymptotically optimal via theoretical analysis, and is further developed to accommodate the multiuser management. The simulation results demonstrate that our proposed algorithm can significantly improve the energy and reliability performance.

Mobile-edge computing (MEC) enhances the capacities and features of mobile devices by offloading computation-intensive tasks over wireless networks to edge servers. One challenge faced by the deployment of MEC in cellular networks is to support user mobility. As a result, offloaded tasks can be seamlessly migrated between base stations (BSs) without compromising the resource-utilization efficiency and link reliability. In this paper, we tackle the challenge by optimizing the policy for migration/handover between BSs by jointly managing computation-and-radio resources. The objectives are twofold: maximizing the sum offloading rate, quantifying MEC throughput, and minimizing the migration cost. The policy design is formulated as a decision-optimization problem that accounts for virtualization, I/O interference between virtual machines (VMs), and wireless multi-access. To solve the complex combinatorial problem, we develop an efficient relaxation-and-rounding based solution approach. The approach relies on an optimal iterative algorithm for solving the integer-relaxed problem and a novel integer-recovery design. The latter outperforms the traditional rounding method by exploiting the derived problem properties and applying matching theory. In addition, we also consider the design for a special case of “hotspot mitigation”, referring to alleviating an overloaded server/BS by migrating its load to the nearby idle servers/BSs. From simulation results, we observed close-to-optimal performance of the proposed migration policies under various settings. This demonstrates their efficiency in computation-and-radio resource management for joint service migration and BS handover in multi-cell MEC networks.

Mobile-edge computing (MEC) enhances the capacities and features of mobile devices via offloading computation-intensive tasks over wireless networks to the edge servers. One challenge faced by the deployment of MEC in cellular networks is to support user mobility, so that the offloaded tasks can be seamlessly migrated between base stations (BSs) without compromising the resource-utilization efficiency and link reliability. In this paper, we tackle the challenge by optimizing the policy for migration/handover between BSs by jointly managing computation-and-radio resources. The policy design is formulated as a multi-objective optimization problem that maximizes the sum offloading rate, quantifying MEC throughput, and minimizes the migration cost, where the issues of virtualization, I/O interference between virtual machines (VMs), and wireless multi-access are taken into account. To solve the complex combinatorial problem, we develop an efficient relaxation-and-rounding based approach, including an optimal iterative algorithm for solving the integer-relaxed problem and a novel integer-recovery design that exploits the derived problem properties. The simulation results show the close-to-optimal performance of the proposed migration policies under various settings, validating their efficiency in computation-and-radio resource management for joint service migration and BS handover in multi-cell MEC networks.

Exploiting the idle computation resources of mobile devices in mobile edge computing (MEC) system can achieve both channel diversity and computing diversity as mobile devices can offload their computation tasks to nearby mobile devices in addition to MEC server embedded access point (AP). In this paper, we propose a non-orthogonal multiple-access (NOMA)-aided cooperative computing scheme in a basic three-node MEC system consisting of a user, a helper, and an AP. In particular, we assume that the user can simultaneously offload data to the helper and the AP using NOMA, while the helper can locally compute data and offload data to the AP at the same time. We study two optimization problems, energy consumption minimization and offloading data maximization, by joint communication and computation resource allocation of the user and helper. We find the optimal solutions for the two non-convex problems by some proper mathematical methods. Simulation results are presented to demonstrate the effectiveness of the proposed schemes. Some useful insights are provided for practical designs.

Mobile-edge computing (MEC) is an emerging technology for enhancing the computational capabilities of mobile devices and reducing their energy consumption via offloading complex computation tasks to the nearby servers. Multiuser MEC at servers is widely realized via parallel computing based on virtualization. Due to finite shared I/O resources, interference between virtual machines (VMs), called I/O interference, arises that degrades the computation performance. In this paper, we study the problem of joint radio-and-computation resource allocation (RCRA) in multiuser MEC systems in the presence of I/O interference. Specifically, we formulate a sum offloading rate maximization problem by joint offloading-user scheduling, the offloaded size control, and time allocation for communication (offloading and downloading) and computation. The problem is a non-convex mixed-integer programming problem. An optimal algorithm with low complexity is designed based on a decomposition approach and Dinkelbach method. The simulation results demonstrate considering of I/O interference can endow on an offloading controller robustness against the performancedegradation factor.

In this paper, we study energy-efficient resource allocation in distributed antenna system with wireless power transfer, where time-division multiple access is adopted for downlink multiuser information transmission. In particular, when a user is scheduled to receive information, other users harvest energy at the same time using the same radio-frequency signal. We consider two types of energy efficiency (EE) metrics: user-centric EE (UC-EE) and network-centric EE (NC-EE). Our goal is to maximize the UC-EE and NC-EE, respectively, by optimizing the transmission time and power subject to the energy harvesting requirements of the users. For both UC-EE and NC-EE maximization problems, we transform the nonconvex problems into equivalently tractable problems by using suitable mathematical tools and then develop iterative algorithms to find the globally optimal solutions. Simulation results demonstrate the superiority of the proposed methods compared with the benchmark schemes.

Mobile-edge computing (MEC) is an emerging technology for enhancing the computational capabilities of the mobile devices and reducing their energy consumption via offloading complex computation tasks to the nearby servers. Multiuser MEC at servers is widely realized via parallel computing based on virtualization. Due to finite shared I/O resources, interference between virtual machines (VMs), called I/O interference, degrades the computation performance. In this paper, we study the problem of joint radio-and-computation resource allocation (RCRA) in multiuser MEC systems in the presence of I/O interference. Specifically, offloading scheduling algorithms is designed targeting two system performance metrics: sum offloading rate maximization and sum mobile energy consumption minimization. Their designs are formulated as non-convex mixed-integer programming problems, which account for latency due to offloading, result downloading, and parallel computing. A set of low-complexity algorithms are designed based on a decomposition approach and leveraging classic techniques from combinatorial optimization. The resultant algorithms jointly schedule offloading users, control their offloading sizes, and divide time for communication (offloading and downloading) and computation. They are either optimal or can achieve close-to-optimality as shown by simulation. The comprehensive simulation results demonstrate that considering of I/O interference can endow on an offloading controller robustness against the performance-degradation factor.

This paper studies joint uplink (UL) and downlink (DL) resource allocation in an orthogonal frequency-division multiple-access (OFDMA) based cloud radio access network (CRAN), where the distributed remote radio heads (RRHs) cooperatively serve the users' UL and DL transmission over different subcarriers (SCs). We aim to maximize the system throughput through jointly optimizing UL/DL scheduling, SC assignment, RRH selection and power allocation under the maximum power and fronthaul capacity constraints. We formulate the problem as a mixed integer programming problem which is nonconvex and NP-hard. We propose an optimal algorithm based on the Lagrange duality method to solve this problem. We also propose a suboptimal algorithm to further reduce the complexity. Simulation results illustrate that the proposed algorithms can considerably improve the system throughput compared to other benchmark schemes.

The rapid growth of Internet of Things (IoT) dramatically increases power consumption of wireless devices. Simultaneous wireless information and power transfer (SWIPT) is a promising solution for sustainable operation of IoT devices. In this paper, we study energy efficiency (EE) in SWIPT-based distributed antenna system, where power splitting (PS) is applied at IoT devices to coordinate the energy harvesting and information decoding processes by varying transmit power of distributed antenna ports and PS ratios of IoT devices. In the case of single IoT device, we find the optimal closed-form solution by deriving some useful properties based on Karush-Kuhn-Tucker conditions and the solution is no need for numerical iterations. For the case of multiple IoT devices, we propose an efficient suboptimal algorithm to solve the EE maximization problem. Simulation results show that the proposed schemes achieve better EE performance compared with other benchmark schemes in both single and multiple IoT devices cases.

This paper studies a relay-assisted wireless-powered communication network consisting of multiple source-destination pairs and a hybrid relay node (HRN). We consider a “charge-then-forward” protocol at the HRN, in which the HRN with constant energy supply first acts as an energy transmitter to charge the sources, and then forwards the information from the sources to their destinations through time division multiple access (TDMA) or frequency division multiple access (FDMA). Processing costs at the wireless-powered sources are taken into account. Our goal is to maximize the sum rate of all transmission pairs by jointly optimizing the time, frequency, and power resources. The formulated optimization problems for both TDMA and FDMA are non-convex. For the TDMA scheme, by appropriate transformation, the problem is reformulated as a convex problem and be optimally solved. For the FDMA case, we find the asymptotically optimal solution in the dual domain. Furthermore, suboptimal algorithms are proposed for both schemes to tradeoff the complexity and performance. Finally, the simulation results validate the effectiveness of the proposed schemes.

This paper studies the design and secrecy performance of linear multihop networks, in the presence of randomly distributed eavesdroppers in a large-scale 2-D space. Depending on whether there is feedback from the receiver to the transmitter, we study two transmission schemes: an ON–OFF transmission (OFT) and a non-ON–OFF transmission (NOFT). In the OFT scheme, transmission is suspended if the instantaneous received signal-to-noise ratio (SNR) falls below a given threshold, whereas, there is no suspension of transmission in the NOFT scheme. We investigate the optimal design of the linear multiple network in terms of the optimal rate parameters of the wiretap code as well as the optimal number of hops. These design parameters are highly interrelated, since more hops reduce the distance of per-hop communication, which completely changes the optimal design of the wiretap coding rates. Despite the analytical difficulty, we are able to characterize the optimal designs and the resulting secure transmission throughput in mathematically tractable forms in the high SNR regime. Our numerical results demonstrate that our analytical results obtained in the high SNR regime are accurate at practical SNR values. Hence, these results provide useful guidelines for designing linear multihop networks with targeted physical layer security performance.

In this paper, we study the secure beamforming design for a cognitive radio network (CRN), where a primary transmitter-receiver pair coexists with an untrusted secondary transmitter-receiver pair. Each pair constitutes a multiple-input single-output link. We consider an underlay scheme and a cooperative scheme. For the underlay scheme, the secondary user (SU) is allowed to transmit simultaneously in the presence of the primary transmission. For the cooperative scheme, the secondary transmitter acts as a relay to forward the secrecy information of the primary transmission in exchange for its own transmission. For both schemes, the SU is untrusted and considered a potential eavesdropper. Our goal is to minimize the total power consumption while satisfying the primary user's required secrecy rate and the SU's required information rate. Using suitable optimization tools, we design the jointly optimal secure beamforming for the underlay scheme and an alternative optimizing algorithm for the cooperative scheme. To further reduce the complexity, we also design suboptimal zero-forcing beamformers for both schemes. The simulation results verify the proposed schemes.

This paper studies resource allocation in a user- centric cloud radio access network (CRAN), where remote radio heads (RRHs) communicate with users using orthogonal frequency-division multiple-access (OFDMA). We aim to maximize the system throughput through joint RRH selection and SC allocation under fronthaul capacity constraints. We formulate the problem as a non-convex problem which is NP-hard. To solve this complex problem efficiently, we propose a near-optimal algorithm based on gradient method to achieve a suboptimal solution. Simulation results show that the proposed joint resource allocation scheme can attain significant throughput gains compared to other benchmark schemes.

In this paper, we consider “harvest-then-use” based simultaneous wireless information and power transfer (SWIPT) for cooperative relay transmission, where a multiple-input multiple-output (MIMO) relay adopts decode-and-forward (DF) relaying strategy for information forwarding. By considering practical power splitting (PS) relay receiver architecture, joint optimization problem of power allocation and determining power splitting ratios is formulated to maximize the end-to-end achievable rate. Using the Lagrange dual method, we develop the efficient algorithm to find the optimal solution. Several valuable insights are provided via theoretical analysis and simulation results.

The concept of device-to-device (D2D) communications underlaying cellular networks opens up potential benefits for improving system performance but also brings new challenges, such as interference management. In this paper, we propose a pricing framework for interference management from the D2D users to the cellular system, where the base station (BS) protects itself (or its serving cellular users) by pricing the cross-tier interference caused from the D2D users. A Stackelberg game is formulated to model the interactions between the BS and D2D users. Specifically, the BS sets prices to maximize its revenue (or any desired utility) subject to an interference temperature constraint. For given prices, the D2D users competitively adapt their power allocation strategies for individual utility maximization. We first analyze the competition among the D2D users by noncooperative game theory and an iterative-based distributed power allocation algorithm is proposed. Then, depending on how much network information the BS knows, we develop two optimal algorithms, one for uniform pricing with limited network information and the other for differentiated pricing with global network information. The uniform pricing algorithm can be implemented by a fully distributed manner and requires minimum information exchange between the BS and D2D users, and the differentiated pricing algorithm is partially distributed and requires no iteration between the BS and D2D users. Then, a suboptimal differentiated pricing scheme is proposed to reduce complexity and it can be implemented in a fully distributed fashion. Extensive simulations are conducted to verify the proposed framework and algorithms.

In this paper, we enhance physical layer security for downlink heterogeneous networks by using friendly jammers and full-duplex users. The jammers are selected to transmit jamming signal if their interfering power on the scheduled users is below a threshold, meanwhile the scheduled users confound the eavesdroppers using artificial noise by full-duplexing. Using the tools of stochastic geometry, we derive the expressions of connection probability and secrecy probability. In particular, the locations of active jammers are modeled by a modified Poisson hole process. Determining the jammer selection threshold is further investigated for connection probability maximization subject to the security constraints. A greedy algorithm is proposed to efficiently solve this problem. The accuracy of the theoretical analysis and the efficiency of the proposed algorithm are evaluated by numerical simulations.

In this paper, we investigate the physical layer security for heterogeneous networks (HetNets), where the confidential message transmitted to each legitimate user can be eavesdropped by multiple eavesdroppers. To enhance the secrecy performance, we introduce friendly jammers and full duplex users in our model. We propose a jammer selection scheme based on average user received jamming power. The jammers whose interference power on the scheduled users less than a threshold are selected to transmit independent artificial noise to confound the passive eavesdroppers. Furthermore, the scheduled users try to jam the eavesdroppers around using full duplex capacity. Using the tools from stochastic geometry, we derive the theoretical analyses of user connection probability and secrecy probability. In particular, the locations of selected jammers are modeled by a Poisson hole process instead of more popular homogeneous Poisson point process. Our analytical results show a good agreement with the simulations. The ergodic secrecy rate in our scheme outperforms that in the traditional HetNets.

In this paper, we consider the secure connection problem in multihop wireless networks with full-duplex (FD) jamming relaying, where the colluding eavesdroppers are randomly distributed following a homogeneous Poisson point process (PPP). By applying FD, each legitimate node (including relay and destination) jams the eavesdroppers when it receives the desired signal from transmitter. We adopt the end- to- end secure connection probability (SCP) as a secrecy metric to characterize the physical layer security performance. We first derive the exact expression of SCP for any given path. Then, an approximation of the SCP is proposed to facilitate efficient secure routing by using a revised Bellman- Ford algorithm. We show that a notable performance gain can be achieved by the proposed scheme compared to the half-duplex (HD) scheme, if the self- interference can be well canceled. Simulation results verify our theoretical analysis.

In this paper, we study the problem of secure routing in a multihop wireless ad-hoc network in the presence of randomly distributed eavesdroppers. Specifically, the locations of the eavesdroppers are modeled as a homogeneous Poisson point process (PPP) and the source-destination pair is assisted by intermediate relays using the decode-and-forward (DF) strategy. We analytically characterize the physical layer security performance of any chosen multihop path using the end-to-end secure connection probability (SCP) for both colluding and noncolluding eavesdroppers. To facilitate finding an efficient solution to secure routing, we derive accurate approximations of the SCP. Based on the SCP approximations, we study the secure routing problem, which is defined as finding the multihop path having the highest SCP. A revised Bellman–Ford algorithm is adopted to find the optimal path in a distributed manner. Simulation results demonstrate that the proposed secure routing scheme achieves nearly the same performance as exhaustive search.

In this paper, we study the simultaneous wireless information and power transfer in downlink multiuser orthogonal frequency-division multiple access systems, where each user applies power splitting to coordinate the energy harvesting and secrecy information decoding processes. Assuming equal power allocation across subcarriers, we formulate an optimization problem to maximize the aggregate harvested power of all users while satisfying secrecy rate requirements of individual users by joint subcarrier allocation and power splitting ratio selection. Due to the NP-hardness of the problem, we propose two suboptimal algorithms to solve the problem. The first one is an iterative algorithm that optimizes subcarrier allocation and power splitting ratios by an alternating way in dual domain. The second algorithm is based on a two-step approach that allocates subcarriers and selects power splitting ratios sequentially. The numerical results show that the proposed methods outperform the conventional methods and provide good trade offs between performance and complexity.

In this paper, a novel hybrid-duplex scheme based on received power is proposed for heterogeneous networks (HetNets). In the proposed scheme, the duplex mode (half or full duplex) of each user is switchable according to the received power from its serving base station. The signal-to-interference-plus-noise-ratio and spectral efficiency are analyzed for both downlink and uplink channels by using the tools of inhomogeneous Poisson point process. Furthermore, determining power threshold for duplex mode switching is investigated for sum rate maximization, which is formulated as a nonlinear integer programming problem and a greedy algorithm is proposed to solve this problem. The theoretical analysis and the proposed algorithm are evaluated by numerical simulations. Simulation results show that the proposed hybrid-duplex scheme outperforms the half-duplex or full-duplex HetNet schemes.

In this paper, we study simultaneous wireless information and power transfer (SWIPT) in orthogonal frequency division multiple access (OFDMA) systems with the coexistence of information receivers (IRs) and energy receivers (ERs). The IRs are served with best-effort secrecy data and the ERs harvest energy with minimum required harvested power. To enhance the physical layer security for IRs and yet satisfy energy harvesting requirements for ERs, we propose a new frequency-domain artificial noise (AN) aided transmission strategy. With the new strategy, we study the optimal resource allocation for the weighted sum secrecy rate maximization for IRs by power and subcarrier allocation at the transmitter. The studied problem is shown to be a mixed integer programming problem and thus nonconvex, while we propose an efficient algorithm for solving it based on the Lagrange duality method. To further reduce the computational complexity, we also propose a suboptimal algorithm of lower complexity. The simulation results illustrate the effectiveness of proposed algorithms as compared against other heuristic schemes.

In this paper, we propose a novel distance-based hybrid-duplex scheme for heterogeneous networks (HetNets). Assuming each user can select half- or full-duplex mode based on the transmit distance to its serving base station (BS), we derive the bidirectional full-duplex and downlink half-duplex signal-tointerference-ratio (SINR) and achievable rate using stochastic geometry. The numerical results show good a agreement with our analyze and verify that the proposed distance-based hybridduplex scheme significantly outperforms the conventional HetNet schemes.

In heterogeneous networks (HetNets), mobile users are proactively offloaded to small cells by association bias. To alleviate the signal-to-interference-plus-noise-ratio (SINR) degradation to offloaded users, resource partitioning and transmit power reduction are proposed in Long Term Evolution-Advanced (LTE-A), which is known as enhanced inter-cell interference coordination (eICIC). In this paper, we develop a tractable framework for performance analysis of eICIC with joint cell association, resource partitioning, and transmit power reduction by using stochastic geometry. We derive the downlink coverage probability of the network. Numerical results are provided and we give some valuable insights and guidelines for eICIC in cochannel HetNets.

In this paper, we consider a cognitive radio network (CRN) consisting of a primary transmitter-receiver pair and an untrusted secondary transmitter-receiver pair, and each pair is a multiple-input single-output (MISO) link. We consider two transmission schemes, namely underlay scheme and cooperative scheme. For the underlay scheme, the secondary user (SU) is allowed to transmit simultaneously in the presence of primary transmission. For the cooperative scheme, the secondary transmitter acts as a relay node to increase the secrecy rate of primary transmission in exchange for its own transmission. For both schemes, the SU is untrusted and considered as a potential eavesdropper. Our goal is to minimize the total power consumption while satisfying the primary user (PU)'s required secrecy rate and SU's required information rate. By suitable optimization tools, we design the joint secure beamforming for both schemes. The simulation results show that in the considered system model, the underlay scheme performs better than the cooperative scheme, especially with high rate requirements and large number of antennas at secondary transmitter.

In this paper, we consider simultaneous wireless information and power transfer (SWIPT) in orthogonal frequency division multiple access (OFDMA) systems with the coexistence of information receivers (IRs) and energy receivers (ERs). The IRs are served with best- effort secrecy data and the ERs harvest energy with minimum required harvested power. To enhance physical- layer security and yet satisfy energy harvesting requirements, we introduce a new frequency-domain artificial noise based approach. We study the optimal resource allocation for the weighted sum secrecy rate maximization via transmit power and subcarrier allocation. The considered problem is nonconvex, while we propose an efficient algorithm for solving it based on Lagrange duality method. Simulation results illustrate the effectiveness of the proposed algorithm as compared against other heuristic schemes.

In this work, we maximize the secrecy rate of the wireless-powered untrusted amplify-and-forward relay networks by jointly designing power splitting (PS) ratio and relay beamforming with the proposed global optimal algorithm (GOA) and local optimal algorithm (LOA). To guarantee secure communication, the destination-based artificial noise is sent to degrade the reception of the untrusted relay, and it also becomes a new source of energy powering relay to forward the information with power splitting (PS) technique. Simulation result shows that LOA can achieve satisfactory secrecy rate performance compared with that of GOA, but with less computation time. It also manifests that both proposed algorithms outperform the benchmark method.

In this paper, we consider simultaneous wireless information and power transfer (SWIPT) in an orthogonal frequency-division multiplexing (OFDM) relaying system, where a source node transfers information and a fraction of power simultaneously over OFDM to a relay node, and the relay node uses the harvested power from the source node to forward the source information to the destination. To support such simultaneous information and energy cooperation, we propose two transmission protocols, namely power splitting (PS) relaying protocol and the transmission mode adaptation (TMA) protocol for without/with the source-destination link, respectively. For both transmission protocols, joint resource allocation problems are formulated to maximize the system throughput. By using the Lagrange dual method, we develop efficient algorithms to find the optimal polices of the nonconvex optimization problems.

Random wireless networks with finite number of nodes distributed uniformly in a circular-shaped finite region are considered. A closed-form expression for the complimentary cumulative distribution function (CCDF) of Euclidean distance between a randomly selected node and its ith nearest node is derived. In contrast to most of the existing literature, the node observed is not necessarily required to be located at the centre of the network region nor located at fixed point. Moreover, the joint distribution of the distances between the ith and jth nearest nodes is derived. Finally, a tractable and quite tight lower bound on outage performance of the uniform network based on the dominant interferer node is derived. Using our derived lower bound, we can determine the maximum number of nodes that should be deployed to satisfy the required QoS. Computer simulation illustrates the validity of the theoretical analysis.

In heterogeneous networks (HetNets), user association is introduced to encourage more users to connect to lightly loaded small cells. This is known as data offloading. However, this may result in that small cells usually sacrifice their own users' performances since the offloaded macro users consume the resources of small cells. In this paper, we aim to improve the performances of both macrocell and small cells by channel assignment, biasing and power allocation in a distributed manner. The joint problem is modeled as a Stackelberg game. As the leader, the macrocell keeps silent on some channels for reducing interference to motivate the small cells to adapt macro users. As the followers, the small cells operate with biasing to accommodate macro users. Meanwhile, both the macrocell and small cells optimally allocate the transmit power over channels. We propose efficient algorithms for the macrocell and small cells making strategies. Furthermore, we prove that the proposed algorithms can lead to an equilibrium of the game. Simulation results show that the proposed methods significantly improve the network performances, compared with the existing methods.

Interference coordination between high-power macros and low-power picos deeply impacts the performance of heterogeneous networks (HetNets). It should deal with three challenges: user association with macros and picos, the amount of almost blank subframe (ABS) that macros should reserve for picos, and resource block (RB) allocation strategy in each eNB. We formulate the three issues jointly for sum weighted logarithmic utility maximization while maintaining proportional fairness of users. A class of distributed algorithms are developed to solve the joint optimization problem. Our framework can be deployed for enhanced inter-cell interference coordination (eICIC) in existing LTE-A protocols. Extensive evaluation are performed to verify the effectiveness of our algorithms.

In this paper, we propose a pricing framework for interference management in device-to-device (D2D) underlaying cellular networks, where the base station (BS) protects itself by pricing the cross-tier interference caused from the D2D users. A Stackelberg game is formulated to model the interactions between the BS and D2D users. Specifically, the BS sets prices to a maximize its revenue subject to an interference temperature constraint. For given specified prices, the D2D users competitively adapt power allocation strategies for their individual utility maximization. We first analyze the competition among the D2D users by noncooperative game theory and an iterative based distributed power allocation algorithm is proposed. Then, depending on how much network information the BS knows, we develop two pricing algorithms, i.e., uniform pricing with limited network information and differentiated pricing with global network information.

In this paper, we study the cross layer design and optimization for delay quality-of-service (QoS) provisioning in spectrum sharing femtocell networks. Our goal is to find the optimal resource allocation policy to maximize the throughput for each femtocell user, addressing the co-channel interference problem in the physical layer and the individual statistical delay-QoS guarantee problem from the upper layers. The statistical delay-QoS requirement is characterized by the QoS exponent. By integrating the concept of effective capacity, the cross-layer optimization problem is formulated as an effective capacity maximization game. With partial dual decomposition, this game is solved through a hierarchical structure. Specifically, we derive the optimal power allocation policy for femtocell users and design a distributed algorithm to obtain the Nash Equilibrium (N.E.). Numerical results show that the proposed policy can efficiently improve the performance of the networks.

From security perspective, a friendly relay may help to keep the confidential messages from being eavesdropped, while an untrusted relay may intentionally eavesdrop the messages when relaying. This paper studies the secure beamforming for multiple-input multiple-output (MIMO) two-way communications, where two source nodes exchange information with the help of an untrusted relay node. The relay adopts amplify-and-forward (AF) strategy and acts as both an essential helper and a potential eavesdropper. Our goal is to maximize the secrecy sum rate of the bidirectional links by jointly optimizing the source and relay beamformers. For the two-phase two-way relay scheme, we first derive the optimal structure of the relay beamformer and then propose an iterative algorithm to jointly optimize the source and relay beamformers. Then, a comprehensive study on the asymptotical performance is conducted by letting the source and relay powers approach zero or infinity. In particular, we show that when all powers approach infinity, the two-way relay scheme achieves the maximum secrecy rate if the transceiver beamformers are designed such that the received signals at the relay can be aligned to be parallel.

As an effective solution for indoor coverage and service offloading from the conventional cellular networks, femtocells have attracted a lot of attention in recent years. From the perspective of spectral efficiency, the macrocell base station (MBS) and femtocell base stations (FBSs) are usually deployed in the same spectrum. Then the interference problem has become a key obstruction that limits the network performance. In this paper, we study the spectrum reuse in the two-tier LTE femtocell network. In order to improve the network performance, the FBSs are encouraged to provide services to nearby macrocell users, and the MBS releases a fractional spectrum to the FBSs for avoiding cross-tier interference in return. We model this problem as a Stackelberg game where the MBS acts as a leader and the FBSs as the followers. We define the utilities for the MBS and FBSs as the average throughput and the distortion-rate function, respectively. It is worth noting that in our Stackelberg game model, there is no monetary price for the interaction between the leader and followers, which is the significant distinction from previous works. The optimal strategies of spectrum reuse for both MBS and FBSs are proposed by analyzing the Stackelberg game model. The simulation results show that the proposed spectrum reuse scheme can significantly improve the network performance.

Most existing works about scheduling and resource allocation for orthogonal frequency division multiplexing (OFDM) based two-way relay networks have focused on immediate relay forwarding. In this paper, we consider relay buffering in delay-tolerant networks. The relay node is aided by two buffers and one for each user, so that it can adaptively decide when to buffer the received packets or to forward them according to the instantaneous channel and queue conditions. We formulate the joint optimization of subcarrier assignment, transmission mode selection (direct or relay mode), and relay strategy selection (buffering or forwarding), for maximizing the long-term average throughput. An efficient dual-based algorithm is proposed to characterize the optimal policy. Simulation results show that relay buffering can significantly enhance the long-term throughput in OFDM bidirectional transmission systems.

Two-way relaying promises considerable improvements on spectral efficiency in wireless relay networks. While most existing works focus on physical layer approaches to exploit its capacity gain, the benefits of two-way relaying on upper layers are much less investigated. In this paper, we study the cross-layer design and optimization for delay quality-of-service (QoS) provisioning in two-way relay systems. Our goal is to find the optimal transmission policy to maximize the weighted sum throughput of the two users in the physical layer while guaranteeing the individual statistical delay-QoS requirement for each user in the datalink layer. This statistical delay-QoS requirement is characterized by the QoS exponent. By integrating the concept of effective capacity, the cross-layer optimization problem is equivalent to a weighted sum effective capacity maximization problem. We derive the jointly optimal power and rate adaptation policies for both three-phase and two-phase two-way relay protocols. Numerical results show that the proposed adaptive transmission policies can efficiently provide QoS guarantees and improve the performance. In addition, the throughput gain obtained by the considered three-phase and two-phase protocols over direct transmission is significant when the delay-QoS requirements are loose, but the gain diminishes at tight delay requirements. It is also found that, in the two-phase protocol, the relay node should be placed closer to the source with more stringent delay requirement.

Two-way relaying can considerably improve spectral efficiency in relay-assisted bidirectional communications. However, the benefits and flexible structure of orthogonal frequency division multiplexing (OFDM)-based two-way relay systems is much less exploited. Moreover, most of existing works have not considered quality-of-service (QoS) provisioning for two-way relaying. In this paper, we consider the OFDM-based bidirectional transmission where a pair of users exchange information via the assistance of a decode-and-forward (DF) relay. Each user can communicate with the other via three transmission modes: direct transmission, one-way relaying, and two-way relaying. We jointly optimize the transmission policies, including power allocation, transmission mode selection, and subcarrier assignment in order to maximize the weighted sum rates of the two users with diverse QoS guarantees. This is formulated as a mixed integer programming problem. By using the dual method, we efficiently solve the problem in an asymptotically optimal manner. Simulation results show that the proposed resource allocation scheme can substantially improve system performance compared with conventional schemes. A number of interesting insights are also obtained via comprehensive simulations.

This paper studies a wireless network where multiple users cooperate with each other to improve the overall network performance. Our goal is to design an optimal distributed power allocation algorithm that enables user cooperation, in particular, to guide each user on the decision of transmission mode selection and relay selection. Our algorithm has the nice interpretation of an auction mechanism with multiple auctioneers and multiple bidders. Specifically, in our proposed framework, each user acts as both an auctioneer (seller) and a bidder (buyer). Each auctioneer determines its trading price and allocates power to bidders, and each bidder chooses the demand from each auctioneer. By following the proposed distributed algorithm, each user determines how much power to reserve for its own transmission, how much power to purchase from other users, and how much power to contribute for relaying the signals of others. We derive the optimal bidding and pricing strategies that maximize the weighted sum rates of the users. Extensive simulations are carried out to verify our proposed approach.

In wireless cooperative orthogonal frequency-division multiple-access (OFDMA) networks, it is important to adapt the transmission strategies for each user according to the network channel dynamics in order to optimize the overall system performance. The adaption involves transmission mode selection (a user can choose from direct or cooperative transmission), subcarrier assignment, subcarrier pairing (the incoming and outgoing subcarriers at the relay for cooperative transmission need to be matched), relay selection, as well as power allocation and hence is highly challenging. Many previous works only consider a subset of the adaptation. In this paper, we tackle the joint optimization problem using a network flow approach. Specifically, we first show that for given power allocation, the combinatorial optimization problem of transmission mode selection, subcarrier assignment, relay selection and subcarrier pairing for the system total throughput maximization can be transformed into a minimum cost network flow (MCNF) problem with integer solutions. The linear optimal distribution (LOD) algorithm is applied to find the optimal solution in polynomial time. We then solve the mixed integer programming problem of the joint assignment and power allocation in an asymptotically optimal way in the dual domain. Simulation results show that the proposed algorithms can significantly enhance the overall system throughput.

In this paper, we consider the orthogonal frequency division multiplexing (OFDM)-based bidirectional transmission where a pair of users exchange information with the assistance of a decode-and-forward (DF) relay. Each user can communicate with the other via three transmission modes: direct transmission, one-way relaying, and two-way relaying. We jointly optimize the transmission policies, including power allocation, transmission mode selection, and subcarrier-node assignment for maximizing the weighted sum rates of the two users with quality-of-service (QoS) guarantees. We formulate the joint optimization problem as a mixed integer programming problem. By using the dual method, we solve the problem efficiently in an asymptotically optimal manner. Particularly, we derive the capacity region of two-way DF relaying in parallel relay channels. Simulation results show that the proposed resource-allocation scheme can substantially improve system performance compared with the conventional schemes.

In this paper, we consider the cross-layer design for delay quality-of-service (QoS) provisioning in two-way relay systems. We aim to find the optimal resource allocation policy to maximize the weighted sum-rate while guaranteeing the statistical delay-QoS requirements for both users. The delay requirement is characterized as the QoS exponent. With the integration of the concept of effective capacity, the cross-layer optimization problem is equivalent to a weighted sum effective capacity maximization problem. We derive the optimal joint power and rate adaptation policy for the two-phase two-way relaying. Numerical results show that the proposed policy can efficiently support diverse QoS requirements and significantly improve the performance compared with both the fixed power scheme and the weight-based method.

This paper considers a relay-assisted cooperative network where multiple relays assist the communication of multiple users using orthogonal frequency-division multiple-access (OFDMA). Our goal is to improve system performance by exploring full potential of the network in various dimensions, including user, relay, channel, and transmission mode. We formulate the joint optimization of subcarrier pairing, subcarrier assignment, relay selection, and transmission mode selection. We show that this combinatorial optimization problem can be transformed into a minimum cost network flow (MCNF) problem with integer solutions in graph theory. Then the linear optimal distribution (LOD) algorithm is applied to find the optimal solution in polynomial time. Simulations show that the proposed algorithm can significantly enhance the overall system throughput.

Efficient utilization of radio resources in wireless networks is crucial and has been investigated extensively. This letter considers a wireless relay network where multiple user pairs conduct bidirectional communications via multiple relays based on orthogonal frequency-division multiplexing (OFDM) transmission. The joint optimization of channel and relay assignment, including subcarrier pairing, subcarrier allocation as well as relay selection, for total throughput maximization is formulated as a combinatorial optimization problem. Using a graph theoretical approach, we solve the problem optimally in polynomial time by transforming it into a maximum weighted bipartite matching (MWBM) problem. Simulation studies are carried out to evaluate the network total throughput versus transmit power per node and the number of relay nodes.

NA

This paper considers a relay-assisted bidirectional cellular network where the base station (BS) communicates with each mobile station (MS) using orthogonal frequency-division multiple-access (OFDMA) for both uplink and downlink. We first introduce a novel three-time-slot time-division duplexing (TDD) transmission protocol. This protocol unifies the direct transmission, one-way relaying and network-coded two-way relaying between the BS and each MS. Using the proposed TDD protocol, we then propose an optimization framework for resource allocation to achieve the following gains: cooperative diversity gain (via relay selection), network coding gain (via bidirectional transmission mode selection), and multiuser diversity gain (via subcarrier assignment). We formulate the problem as an integer programming problem. By establishing its equivalence with a maximum weighted clique problem (MWCP) in graph theory, we show that the problem can be solved using an ant colony optimization (ACO) based metaheuristic algorithm in polynomial time. Simulation results demonstrate that the proposed protocol together with the ACO algorithm significantly enhances the system total throughput compared with conventional schemes.

This paper considers a relay-assisted bidirectional cellular network where the base station (BS) communicates with each mobile station (MS) using orthogonal frequency-division multiple-access (OFDMA) for both uplink and downlink. The goal is to improve the overall system performance by exploring the full potential of the network in various dimensions including user, subcarrier, relay, and bidirectional traffic. In this work, we first introduce a novel three-time-slot time-division duplexing (TDD) transmission protocol. This protocol unifies direct transmission, one-way relaying and network-coded two-way relaying between the BS and each MS. Using the proposed three-time-slot TDD protocol, we then propose an optimization framework for resource allocation to achieve the following gains: cooperative diversity (via relay selection), network coding gain (via bidirectional transmission mode selection), and multiuser diversity (via subcarrier assignment). We formulate the problem as a combinatorial optimization problem, which is NP-complete. To make it more tractable, we adopt a graph-based approach. We first establish the equivalence between the original problem and a maximum weighted clique problem (MWCP) in graph theory. A metaheuristic algorithm based on ant colony optimization (ACO) is then employed to find the solution in polynomial time. Simulation results demonstrate that the proposed protocol together with the ACO algorithm significantly enhances the system total throughput.